1. Ahmet Ozbay, Wei Tian and Hui Hu
An Experimental Study on the Effects of Wake Interference on the Performance of Wind Turbines over Flat and Complex Terrains
Advanced Flow Diagnostics and Experimental Aerodynamics Laboratory Department of Aerospace Engineering
Ahmet Ozbay, Wei Tian and Hui Hu
Iowa State University Howe Hall, Ames, Iowa

2. OUTLINE Motivation/Objectives Introduction
Experimental set-up and procedure
Investigation of wake interference effects
Wake interference of wind turbines with different spacing
Wake interference within an array of turbines in a line
Multiple wake interactions in wind farms with different layouts
Upstream turbine operating(yaw) conditions
Terrain (2D-Ridge) effects on the wake interferences among multiple wind turbines
Summary

3. Objectives of the Present Study
The most important aerodynamic aspects in the design of wind farms:
Velocity deficit:
The velocity deficit is linked to the amount of power can be extracted from the flow.
The power output could lose up to 40% when the wind turbine placed in the wake.
Enhancement of turbulence intensity:
Enhanced turbulence intensity in the wake is associated directly with the fatigue loads and failure of the wind turbine components.
(Barthelmie et al., 2007)
How to reduce the wake induced effects over flat and complex terrains?
(1) Investigating the effects of the array spacing and wind turbines layout on the wake interference and performance of multiple wind turbines sited in a wind farm for higher total power yield and better durability.
(2) Investigating the effects of topography (complex terrains -2D Ridge) on the wind turbine performance as well as on the wake interaction
(3) Investigating the effects of upstream turbine operating (yaw, pitch) conditions on the efficiency of wind farm

4. Objectives of the Present Study
Comparison of onshore and offshore winds:
Offshore wind farms:
Wind turbines sitting on flat ocean surface
Near neutral atmospheric boundary layer winds
High wind speed with relatively low ambient turbulence level
Suffers from ‘deep array effect’
Onshore wind farms:
Wind turbines sitting over complex terrains.
Atmospheric stability is rarely close to near-neutral (highly convective – unstable during the day time and highly stable nocturnal conditions with high shear at night time)
Much higher ambient turbulence level
(4) Different characteristics of atmospheric boundary layer winds were simulated to compare the performances of wind turbines sited in onshore and offshore wind farms.

5. INTRODUCTION – BOUNDARY LAYER
LOG-WIND PROFILE
POWER LAW PROFILE

6. INTRODUCTION – WIND POWER
POWER AVAILABLE IN THE WIND
WIND POWER DENSITY (WPD)(watts/m2) is used to classify the winds
Wind speed plays a crucial role in the wind power
10% increase in the wind speed leads to 33% increase in the wind power density
Wind resource assessment is important for wind turbine siting

7. INTRODUCTION – WIND RESOURCE ASSESSMENT
WIND MEASUREMENT TOOLS
METEOROLOGICAL TOWERS
ANEMOMETERS ( WIND SPEED)
WIND VANES ( WIND DIRECTION)
SENSORS (TEMPERATURE, PRESSURE)
Weibull /Rayleigh probability function
Using probability density function over a wide range of wind speed to estimate the mean power from a turbine

8. INTRODUCTION – WIND RESOURCE ASSESSMENT (2)
Weibull /Rayleigh probability function
Shape factor , k , shape of the curve depending on the standard deviation of the wind speed (σu)
As k increases, mean velocity tends to increase and wind speed variations (σu) fall down

9. INTRODUCTION – WIND TURBINE POWER
Limitations for wind turbine power
Betz limit – theoretical limit (Cp=0.59)
Cut in and cut out speeds
Power losses (wake, environmental, electrical, etc.)
Optimum range
( )

10. INTRODUCTION – WIND ENERGY IN U.S.
One of the fastest growth in terms of electric resource capacity, in GW, every year since 2005
US policy suggest that wind energy will continue to play dominant role in needs of new electric resources in the world, US, Midwest, & Iowa
Wind turbine technology is in its infancy – needs to develop along multiple dimensions over the next 40 years
According to Department of Energy (DOE) recent report, US wind power can reach 300GW by 2030, with on-shore (land-based) wind capacity being a major contributor
Other predictions suggest as much as 600 GW by 2035

11. INTRODUCTION – WIND ENERGY IN U.S.
A target of 20% of US electricity from wind energy by 2030 has been set up by the U.S. Department of Energy (DOE).
Iowa is second in the nation in installed wind energy capacity and it has the highest density of wind power generation capacity with 29.9 kW/km2
According to the Energy Information Administration (EIA), Iowa has reached the milestone of 20% of the state’s electricity, supplying the state with a full one fifth of its energy needs.
Top Wind Energy Production States:
Texas: 10,377 MW
Iowa: 4,322 MW
California: 3,927 MW
Illinois: 2,743 MW
Minnesota: 2,733 MW
Washington: 2,573 MW
Oregon: 2,513 MW
(The data as of Feb 28, 2012)

12. EXPERIMENTAL SET-UP AND PROCEDURE
Parameter R
(mm)
H (mm)
d pole
d nacelle 
(deg.) a a1 a2
Dimension
127
225 18 5o 78 15 50
Cobra probe
JR3 Force/Moment Transducer
Measured parameters:
Dynamic wind loads
Power output and rotational frequency of wind turbine models
Detailed flow field (mean velocity and turbulence) measurements with cobra probe
1:350 scaled model to simulate a 2MW wind turbine with 90m rotor blades
Optical tachometer

13. EXPERIMENTAL SET-UP AND PROCEDURE
127 mm
ERS-100 prototype of wind turbine blade developed by TPI

14. Simulation of Incoming Flow with Different Turbulence Levels
Terrain Category
Terrain description
Gradient height, ZG (m)
Roughness length, ZO (m)
Wind Speed exponent,  1
Open sea, ice, tundra desert
250
0.001
0.11 2
Open country with low scrub or scattered trees
300
0.03
0.15 3
Suburban area, small towns, well wooded areas
400
0.3
0.25 4
Tall buildings, city centers, well developed industrial areas
500
3.0
0.36
Wind speed profile of in atmospheric boundary layer (ABL):
POWER LAW PROFILE
Onshore wind farm
 =0.15
Offshore wind farm
 =0.11
Low turbulence intensity case
(10% at hub height)
High turbulence intensity case
(18% at hub height)

15. Simulation of Incoming Flow with Different Turbulence Levels
Low turbulence intensity case
(10% at hub height)
High turbulence intensity case
(18% at hub height)
Open Terrain Category
Terrain description
Wind Speed exponent, 
Mean wind speed at the hub height, Um
Standard deviation at the hub height wind speed, σ
Variance at the hub height wind speed, σ2
Iuu
Turbulence Intensity (%)
(σ/Um) 1
Open sea, ice, tundra desert
0.11
5.35 m/s
0.56
0.32 10 2
Open country with low scrub or scattered trees
0.15
4.86 m/s
0.88
0.78 18

16. Simulation of Incoming Flow with Different Turbulence Levels
Wind speed distribution at the hub height
Weibull dimensionless shape factor (k) – breadth of the wind speed distribution
The variation of the shape parameter with the incoming flow turbulence (σ/Um)
As the incoming flow turbulence level decreases, the shape of the distribution tends to be tight – less variation in the wind speed (offshore)

17. Investigation of wake interference effects – A
Effect of spacing

18. Investigation of wake interference effects – A
Wake effects – PIV results
Phase locked PIV results
Ensemble averaged (free-run)PIV results

19. Investigation of wake interference effects – A
Wake effects – PIV results
Low turbulence
High turbulence
The evolution of the wake vortex structure (phase-locked measurements)
Power spectrum of the velocity fluctuations (u’) at the top-tip height (x/D=0.5)

20. Investigation of wake interference effects – B 5 Turbines in a line
Low turbulence intensity case
(10% at hub height)
High turbulence intensity case
(18% at hub height) 6D
pos 1
pos 2
pos 3
pos 4
pos 5
(Barthelmie et al., 2007)

21. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts
(a) aligned wind farm with stream-wise spacing 3D
(b) staggered wind farm with stream-wise spacing 3D
(c) aligned wind farm with stream-wise spacing 6D

22. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts

23. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts

24. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts

25. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts

26. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts
Normalized power output
(P/Palone )
Alone
Aligned 3D-spacing
Staggered 3D-spacing
Aligned 6D-spacing
Low turbulence inflow
1.00
0.42
0.73
0.71
High turbulence inflow
0.57
0.85
0.78

27. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts
Wind turbine position
Alone
Aligned 3D-spacing
Staggered 3D-spacing
Aligned 6D-spacing
Low Turbulence Inflow
Thrust Coefficient CT
0.405
0.233
0.321
0.320
Low Turbulence Inflow
Bending moment Coefficient CMy
0.467
0.255
0.364
0.356
High Turbulence Inflow
Thrust Coefficient CT
0.404
0.312
0.388
0.382
0.469
0.348
0.433
0.427

28. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts

29. 0.222 0.145 0.145 Standard Deviation of Thrust Force
Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts
Aligned 3D-spacing
Staggered 3D-spacing
Aligned 6D-spacing
0.222
0.145
0.145
Standard Deviation of Thrust Force

30. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts

31. 0.277 0.201 0.190 Standard Deviation of Thrust Force
Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts
Aligned 3D-spacing
Staggered 3D-spacing
Aligned 6D-spacing
0.277
0.201
0.190
Standard Deviation of Thrust Force

32. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts
Wind farm efficiency :
CP total : The total power output of wind farm
CP alone : Power output of single wind turbine under incoming flow
n : The number of wind turbine in the wind farm

33. Investigation of wake interference effects – C
Multiple wake interactions in wind farms with different layouts
3D Aligned wind farm
In staggered wind farm, the velocity deficit and added turbulence is much lower compared with the aligned wind farm.  
Staggered wind farm is much more efficient than the aligned wind farm with similar stream-wise and span-wise turbine spacing.
The turbulence level of atmospheric boundary layer wind is effective on the wind farm efficiency.
The improvement in the wind farm efficiency due to the incoming flow turbulence is more pronounced in the staggered wind farm.
3D Staggered wind farm

34. Investigation of wake interference effects – D
Upstream Turbine operating (yaw) conditions
Two turbines in tandem arrangement with 2D spacing
Upstream turbine is installed on a turn table and yawed up to 50˚ with an increment of 10˚
Test cases:
Flow measurements in the wake of the upstream turbine with yaw angle from 0˚ to 50˚
The power output and dynamics forces for both upstream and downstream turbine
Flow field measurements in the near wake of the downstream turbine with the yaw angle of upstream turbine changing from 0˚ to 50˚
Experimental set-up for wind tunnel testing 2D g U∞

35. Investigation of wake interference effects – D
Upstream Turbine operating (yaw) conditions
Previous studies on upstream turbine operating (yaw) conditions
Adaramola & Krogstad (2011)
Effect of yaw angle on the performance of upstream turbine
Effect of upstream turbine yaw angle on the downstream turbine performance

36. Investigation of wake interference effects – D
Upstream Turbine operating (yaw) conditions
Previous studies on upstream turbine operating (yaw) conditions
Effect of yaw angle on the wind turbine performance
Pri Mamidipudi (2011) Yaw control: The forgotten controls problem
It was found that there is a cos3 dependency between loss of power and yaw angle especially between -20ᵒ and +20ᵒ for a scaled wind turbine model.

37. Investigation of wake interference effects – D
Upstream Turbine operating (yaw) conditions
Uhub = 5.1 m/s – high turbulence inflow
Uhub = 6.1 m/s – low turbulence inflow 2D g U∞
P (g) ≈ 0.5Cp⍴ AUeff3
P (g) ≈ 0.5Cp ⍴ AU∞3cos3(g)
P (g) ≈ cos3(g)P0_yaw
F(g) ≈ 0.5Cp⍴ AUeff2
F(g) ≈ 0.5Cp⍴ AU∞2cos2(g)
F(g) ≈ cos2(g) F0_yaw

38. Investigation of wake interference effects – D
Upstream Turbine operating (yaw) conditions 2D g U∞

39. Investigation of wake interference effects – D
Upstream Turbine operating (yaw) conditions
Vertical velocity profile at x/D =2 downstream:
Low turbulence 2D g U∞
High turbulence
Wake is deflected sideways by yawing the upstream turbine which results in reduced velocity deficit in the wake.
Effect of yawing the upstream turbine is less pronounced in the wake for the high turbulence flow due to the highly turbulent nature of the flow – turbulent mixing.

40. Investigation of wake interference effects – D
Upstream Turbine operating (yaw) conditions
The overall efficiency of the wind farm (2 turbines):
Effects of upstream turbine yaw angle
Decrease the upstream wind turbine power output with a cos3 dependency between loss of power and yaw angle
Increase the power output of downstream wind turbine
For the low turbulence inflow, the increase of overall power output is up to 6% at an appropriate upstream turbine (α=10˚) yaw angle
For the high turbulent inflow, yawing the upstream turbine does not make any improvement on the overall wind farm power output
Low turbulence
High turbulence

41. Investigation of wake interference effects – E
Complex terrain (2D-Ridge) effects
CASE 1 – Moderate slope (H/L = 0.22), slope = 12˚
CASE 2 – High slope (H/L = 0.41), slope = 22˚
Separation on the lee side (effect of the slope)
(reduced mean speed and higher turbulence levels)
Speed-up effects
(Higher wind speeds, great potential for energy production)
According to Arya (1988), the largest speed-ups are observed over three-dimensional hills of moderate slope.
3D Hills are found to produce lower wind speed increases than 2D Ridges

42. Investigation of wake interference effects – E
Complex terrain (2D-Ridge) effects
Gaussian curve:
h/2 h
Moderate slope 2D-Ridge
Slope =12° 3D 3D 3D 3D
h/2 h
High slope 2D-Ridge
Slope =22° 3D 3D 3D 3D D
Flat surface 3D 3D 3D 3D
pos1
pos2
pos3
pos4
pos5

43. Investigation of wake interference effects – E
Complex terrain (2D-Ridge) effects
Mean velocity and turbulence intensity profile
Wind turbine position
pos1
pos2
pos3
pos4
pos5
Total
Power output flat surface
(normalized with power output of single wind turbine sited on flat surface )
1.00
0.83
0.76
0.75 
0.74
4.08
Power output moderate slope 2D-Ridge
0.91
0.82
1.69
1.02 
0.73
5.17
(~26% more)
Power output high slope 2D-Ridge
0.92
0.63
1.33
0.04 
0.19
3.11
(~24% less)

44. Investigation of wake interference effects – E
Complex terrain (2D-Ridge) effects
Moderate slope 2D-Ridges 6D 6D 6D 6D
Measuring the characteristics of surface winds over complex terrains (hill and valley )
The performances of single wind turbine sited over different locations over complex terrains

45. Investigation of wake interference effects – E
Complex terrain (2D-Ridge) effects
Moderate slope 2D-Ridges 6D 6D 6D 6D
Wind turbine position
pos1
pos2
pos3
pos4
pos5
Power output
(normalized with power output of single wind turbine sited on flat surface )
0.90
1.91
0.67
2.13
0.91
Wind turbine position
pos1
pos2
pos3
pos4
pos5
Two hill
Thrust Coefficient CT
0.117
0.282
0.093
0.298
0.131
Bending moment Coefficient CMZ
0.124
0.258
0.096
0.284
0.130

46. SUMMARY
Factors affecting the complex dynamics of the wind farms were investigated in detail;
Turbine spacing
Wind farm layout (aligned and staggered)
Upstream turbine operating (yaw) conditions
Terrain effects (flat and complex terrain)
Incoming flow character and its interaction with different wind farm layouts

47. Thank you!